race-condition

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes multiple pre-formatted shell commands using curl and complex Python scripts using the httpx and asyncio libraries. These are intended for the agent to execute to perform concurrent network requests against target endpoints.
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection due to how it handles external data:
  • Ingestion points: The agent is instructed to read engagement states from an MCP server and analyze HTTP response headers and bodies from untrusted web servers.
  • Boundary markers: There are no instructions provided to use delimiters or to ignore potential instructions embedded within the data retrieved from target web applications.
  • Capability inventory: The agent has the capability to execute shell commands, run Python scripts, perform network operations, and write files to the local engagement/evidence/ directory.
  • Sanitization: The skill lacks instructions for validating, filtering, or sanitizing external content before it is processed by the agent or used in subsequent logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 09:15 AM
Security Audit — agent-trust-hub — race-condition